Automated Diagnosis of Appendicitis Based on Clinical Notes
Care Process & Redesign
Technology
National Healthcare Innovation and Productivity Medals
National University Health System
31 December 2023
Develop a deep learning AI system for appendicitis diagnosis to assist A&E doctors. AI integration enhanced decision-making for appendicitis diagnosis; scalability for other conditions.
Year Submitted: 2023
Published Date: 31 December 2023
Tags: Technology, Care Process ＆ Redesign, Virtual Reality, Non-Immersive VR, Data Analytics, Artificial Intelligence, Quality Improvement, Job Effectiveness
About this Content
Aims
Develop a deep learning AI system for appendicitis diagnosis to assist A&E doctors.
Background
Manual methods and varying clinician expertise limited diagnostic accuracy and consistency.
Methods
Combined CNN, RNN, and residual networks; evaluated with a randomized clinical trial.
Results
Machine performance improved clinician predictions, especially for junior doctors; F1 scores improved from 0.58 to 0.61 with the tool.
Conclusion
AI integration enhanced decision-making for appendicitis diagnosis; scalability for other conditions.
Lessons Learnt
AI tools support less experienced clinicians; ongoing optimization is key.
Keywords
Appendicitis Diagnosis, AI, Deep Learning, Clinical Decision Support
Innovators' Details
Innovators' Details
Healthcare Cluster(s) | National University Health System |
Organization(s) Involved | National University Health System |
Platform(s) | National Healthcare Innovation and Productivity Medals |
Healthcare Professional Group(s) | Medical |
Applicable Specialty or Discipline | Emergency Medicine, Gastroenterology, Surgery |
Project Lead(s) | Ngiam Kee Yuan |
Project Member(s) |
Connect with this contributor!
Mr. Tan Boon Hoi - boon_hoi_tan@nuhs.edu.sg
Project Attachment
C_323_NUHS_NHIP_2023_Automated_Diagnosis_of_Appendicitis_Based_on.pdf